GA3 Treatment Delays the Deterioration of ‘Shixia’ Longan during the On-Tree Preservation and Room-Temperature Storage and Up-Regulates Antioxidants

Gibberellic acids had been proven to improve the fruit quality and storability by delaying deterioration and maintaining the antioxidant system. In this study, the effect of GA3 spraying at different concentrations (10, 20, and 50 mg L−1) on the quality of on-tree preserved ‘Shixia’ longan was examined. Only 50 mg L−1 GA3 significantly delayed the decline of soluble solids (22.0% higher than the control) and resulted in higher total phenolics content (TPC), total flavonoid content (TFC), and phenylalanine ammonia-lyase activity in pulp at the later stages. The widely targeted metabolome analysis showed that the treatment reprogrammed secondary metabolites and up-regulated many tannins, phenolic acids, and lignans during the on-tree preservation. More importantly, the preharvest 50 mg L−1 GA3 spraying (at 85 and 95 days after flowering) led to significantly delayed pericarp browning and aril breakdown, as well as lower pericarp relative conductivity and mass loss at the later stages of room-temperature storage. The treatment also resulted in higher antioxidants in pulp (vitamin C, phenolics, and reduced glutathione) and pericarp (vitamin C, flavonoids, and phenolics). Therefore, preharvest 50 mg L−1 GA3 spraying is an effective method for maintaining the quality and up-regulating antioxidants of longan fruit during both on-tree preservation and room-temperature storage.


Introduction
Longan (Dimocarpus longan Lour.), which originates from China and belongs to the Longan genus of Sapindaceae, is an important subtropical and globally consumed fruit [1,2]. Sweetness is an important commodity property of mature longan fruit, as its soluble solids content (SSC) usually exceeds 20% [3]. More importantly, except being used as an important specialty fruit, longan fruit has been used as traditional Chinese medicine for thousands of years, due to the abundant bioactive components (e.g., polyphenols, flavonoids, alkaloids, polysaccharides) in pericarp, pulp, and seed [4][5][6]. These components in longan showed control fruits. More importantly, jujube fruit treated with 15 mg L −1 GA 3 and 15 mg L −1 GA 3 plus 1% Parka (containing 7.5% stearic acid, 5% cellulose, and 1% calcium) showed a significantly lower respiration rate and SSC, but higher titratable acidity (TA), VC, and total phenolics [20]. The above reports indicated that the application of GAs had a significant effect on fruit ripening, quality (especially content of antioxidants), and storability, but it was worth noting that the effect of GA 3 application depended on the fruit species, variety, application method, dose, treatment time, and organs.
For longan fruit, the effect of preharvest application of phytohormone on quality, antioxidants, and storability had not been investigated. Treatment with 10 mg L −1 GA 3 at 5 DAA increased the weight, volume, longitudinal and traverse diameter, and the weight of the aril of 'Shixia' longan fruit and led to earlier accumulation of sugars (mainly sucrose, glucose, and fructose), but did not affect the proportion and content of sugars at harvest [21]. Similarly, spraying with 30 mg L −1 GA 3 , ABA, or IAA at 10 DAA did not alter the sugar content of mature longan fruit, but did accelerate the sugar accumulation before ripening stages [22]. However, the effect of GA 3 treatment on quality and antioxidants at harvest, as well as storability of longan fruit, was still not investigated. Sugar receding, i.e., a sharp decrease in SSC in pulp, leads to a fast decline in sweetness and deteriorates the edible quality of on-tree preserved longan fruit [3]. In this study, we reported an effective preharvest GA 3 spraying, through our previous screening assay, to delay the sugar receding of 'Shixia' longan fruit during on-tree stages. Further, the effect of GA 3 treatment on the quality, the content of antioxidants, and enzymatic activity in longan pulp, as well as the storability of longan fruit, was analyzed. A widely targeted metabolome was performed to evaluate the effect of this treatment on the primary and secondary metabolites in longan pulp. These results will benefit the development of new methods to delay sugar receding, maintain the nutrition and quality during on-tree preservation, and improve the storability of longan fruit.

Preharvest Spraying Treatments and Fruit Harvest
The fruits from three longan trees (D. longan cv. Shixia, 15 years old), which had even canopy size and fruit load, were selected for treatments. The spraying experiments were carried out in the two consecutive years (2020 and 2021) in the same commercial orchard in Huidong (Guangdong province, China). In the year 2020, the preharvest spraying of 0 (control), 10, 25, and 50 mg L −1 GA 3 (dissolved in 1 mL ethanol and then diluted with 999 mL ddH 2 O) on fruits was performed five times, at 89, 96, 103, 110, and 120 days after flowering (DAF); the time of full bloom was about 10 April 2020. Three bunches of fruits from each tree were selected for each spraying treatment. About 500 mL solution was used for spraying each bunch of fruits (about 60-80 fruits). For each treatment, 9 fruits from each bunch were randomly selected and harvested before each spraying, at 96, 103, 110, and 120 DAF. Finally, the rest fruits from each bunch were harvested at 130 DAF. The fruits harvested from the same tree were considered as one replication.
In the year 2021, the preharvest spraying of 0 (control), 10, 50, and 150 mg L −1 GA 3 (dissolved in 1 mL ethanol and then diluted with 999 mL ddH 2 O) on fruits was performed five times at 85, 95, 105, and 115 DAF (the time of full bloom was about 10 April 2021). Three bunches of fruits from each tree were randomly selected for each spraying treatment. Another 3 fruit bunches from each tree were selected and subjected to spraying (50 mg L −1 GA 3 or control) 2 times at 85 DAF and 95 DAF, and then the fruits were harvested at 105 DAF and used for the storage experiment. The fruits harvested from the same tree were considered as one replication.

Determination of Total Soluble Solids (TSS) Content, TFC, and TPC
The TSS content (%) was measured by a Brix refractometer (PAL-BX/ACID F5, ATAGO Co., Ltd., Tokyo, Japan). The total flavonoid content (mg g −1 ) and total phenolic content (mg g −1 ) of longan pericarp and pulp were measured according to the previously reported  [23,24]. The ethanolic extract used for determination of TFC and TPC was prepared by homogenizing 0.1 g of frozen pericarp or 0.5 g of frozen pulp, which was extracted using 3 mL 80% ethanol and centrifuged at 5000× g for 5 min. After being extracted three times, the ethanolic extracts was combined in a volumetric flask and adjusted to 10 mL using 80% ethanol [24]. The TFC was determined by the Folin-Ciocalteu method [23], while the TPC was measured using a modified colorimetric method, as previously described [24].
2.3. Measurement of the Activity of Phenylalanine Ammonia-Lyase (PAL) and Peroxidase (POD) The PAL activity was assayed according to the reported method [25]. The frozen pulp or pericarp was ground into powder by liquid nitrogen. Then, 0.1 g pericarp (or 0.25 g pulp) was weighed and mixed with 0.75 mL Tris-HCl buffer (50 mM, pH 8.9), containing 2% polyvinylpyrrolidone (w/v), 2 mM EDTA, and 5 mM β-mercaptoethanol), in a 2 mL tube inside an ice-bath. After being fully vortexed, the sample was subjected to centrifugation at 12,000× g and 4 • C for 30 min. The supernatant was used for determining PAL activity. The reaction system consisted of 0.05 mL supernatant, 0.2 mL 50 mM Tris-HCl buffer, and 0.05 mL 20 mM phenylalanine. The mixture was incubated for 30 min at 37 • C, and then the reaction was stopped by adding 30 µL 6 M HCl. A change of 0.01 in the absorbance at 290 nm per minute was recorded as one unit of PAL enzyme activity. The result was expressed as U g −1 FW.
The POD activity was measured by a change in absorbance at 470 nm caused by production of tetraguaiacol from guaiacol in the presence of H 2 O 2 [26]. The frozen pulp or pericarp was ground into powder by liquid nitrogen. Then, 0.2 g pulp or 0.1 g pericarp was weighed and mixed with 1 mL PBS buffer (50 mM, pH 5.5, containing 2% polyvinylpyrrolidone, w/v) in a tube inside an ice-bath. After being fully vortexed and extracted in a sonicator for 10 min, the sample was subjected to centrifugation (15,000× g) for 20 min at 4 • C. The supernatant was used for determining POD activity. Supernatant volume of 0.03 mL was mixed with 150 µL reaction solution, which was prepared by mixing 3 mL 50 mM PBS buffer, 1 mL 2% H 2 O 2 , and 1 mL 50 mM guaiacol. The reaction mixture was incubated for 10 min at 35 • C, and then the reaction was stopped by adding 60 µL 20% (w/v) trichloroacetic acid. A change of 0.01 in the absorbance per minute was recorded as one unit of POD enzyme activity. The result was expressed as U g −1 FW.

Widely Targeted Metabolome Analysis of On-Tree Preserved Longan Pulp Samples
The frozen pulp was crushed by zirconia beads in a MM 400 mixer mill (Retsch, Shanghai, China) for 1.5 min at 30 Hz. Powdered sample (100 mg) was fully mixed with 1.0 mL 70% aqueous methanol and extracted overnight at 4 • C. Each sample was vortexed three times during the period to increase the extraction efficiency. After being centrifuged at 10,000× g for 10 min, the supernatant was collected and filtered through a Carbon-GCB SPE Cartridge (250 mg, 3 mL, CNWBOND, ANPEL). Then, samples were filtered (SCAA-104, 0.22 µm pore size; ANPEL, http://www.anpel.com.cn//Products.html?id=118 (accessed on 22 July 2021) before LC-MS/MS analysis. Multiple reaction monitoring (MRM) of triple quadrupole mass spectrometry was employed to quantify each metabolite [27]. The parameters of LC-MS/MS analysis and data processing (orthogonal partial least squares discriminant analysis, OPLS-DA [28]) were set according to the previously reported method [2]. Based on the OPLS-DA analysis and significance test, the metabolites with variable importance in projection (VIP) value ≥1 and 1.5 fold-change (FC) (or 2.0-FC) (the Control vs. GA 3 -treated sample, t-test, p-value < 0.05) were identified as the significantly differently accumulated metabolites (DAMs). The cpd_id of each identified metabolite was searched by manually examining its exact mass in the KEGG COMPOUND database (https://www.kegg.jp/kegg/compound/ (accessed on 1 August 2021)). The KEGG enrichment analysis of DAMs was performed by the web tool Metabolites Biological Role (MBROLE) 2.0 [29].

Postharvest Treatment, Storage, and Sampling
More than 500 (control or 50 mg L −1 GA 3 -treated) fruits with no disease and no damage were selected and dipped in 500 mg L −1 prochloraz solution for 2 min. The fruits were then dried at room temperature for 20 min. After that, the fruits were packed into fruit trays (about 25-30 fruits per tray) with 0.01 mm-thick polyethylene films and stored at room temperature (RT, 25 ± 1 • C; 85% relative humidity).
The sampling of the RT-stored fruits was performed at 0, 2, 4, 6, 8, and 10 days after harvest (DAH). The pericarp and pulp of fruits were separated. The inner surface of each pericarp was examined, and then the pericarp browning index as well as the aril breakdown index were evaluated (see Section 2.6). The whole collected pericarp and one half of the collected pulp were immediately frozen in liquid nitrogen, ground, and stored at −80 • C until used. The other half of the pulp was used for juicing and determination of the content of TSS (w/w, %), TA (w/v, %), and VC. Inner pericarp browning index and aril breakdown index were estimated according to a previously reported method [3,7]. Before sampling, the inner pericarp browning index was assessed by measuring the extent of the total browned area (TBNA) on pericarp of fruits from each tray: score = 0, no browning; score = 0.01 to 1.00, TBNA was 0-20%; score = 1.01 to 2.00, TBNA was 20.1-40%; score = 2.01 to 3.00, TBNA was 40.1-60%; score = 3.01 to 4.00, TBNA was 60.1-80%; score = 4.01 to 5.00, TBNA was 80.1-100%. The pericarp browning index of each tray was calculated as follows: Pericarp browning index = ∑ (TBNA score of each fruit)/(total evaluated fruits) (1) The aril breakdown index was assessed by measuring the extent of the total breakdown area (TBDA) on each longan aril using the following scales: score = 0, no breakdown; score = 0.01 to 1.00, TBDA was 0-25%; score = 1.01 to 2.00, TBDA was 25.1-50%; score = 2.01 to 3.00, TBDA was 50.1-75%; score = 3.01 to 4.00, TBDA was 75.1-99.9%; score = 5, TBDA was 100%. The aril breakdown index of each bag was calculated using Formula (2): Aril breakdown index = ∑ (TBDA score of each fruit)/(total evaluated fruits) (2)

Determination of TSS, TA, Vitamin C, and Reduced Glutathione (GSH) in Postharvest Longan Samples
The TSS content (w/w, %) and TA content (w/v, %) were measured by a Brix refractometer (PAL-BX/ACID F5, ATAGO Co., Ltd., Tokyo, Japan). The VC content (in juice, g L −1 ; in pericarp, mg g −1 ) of fruits was determined by an ultraviolet-visible spectrophotometry method. A volume of 10 µL longan juice was diluted with 9900 µL oxalic acid solution (2%, w/v, pH 6.0 adjusted with NaOH) and then the absorbance at 267 nm was measured. The VC content was calculated using a standard curve. The extract used for determination of GSH content was prepared by homogenizing 0.1 g of pericarp powder with 900 µL PBS buffer (50 mM, pH 7.0) or 0.2 g of pulp powder with 800 µL PBS buffer. After being fully vortexed, the sample was subjected to centrifugation at 2500× g and 4 • C for 10 min [30]. The contents of GSH of longan pericarp and pulp were measured based on reaction with DTNB by using a commercialized GSH assay kit (A006-2-1) from Nanjing Jiancheng Bioengineering Institute (www.njjcbio.com/products.asp?id=1532 (accessed on 5 October 2020)). The absorbance of each sample was measured at 405 nm and the GSH content was calculated according to a standard curve.

Statistical Analysis
The results obtained were expressed as the mean ± SE of four randomized replicates. The data were subjected to analysis of variance (ANOVA). The statistical significance between the control and the treated sample was analyzed by paired-samples t-tests, while

GA 3 Treatment Affects TSS, Antioxidants, and Related Enzyme Activities in Longan Fruit during Ripening and On-Tree Preservation
Through a set of GA 3 treatments with different concentrations for 2 years, the results suggested that 50 mg L −1 GA 3 had a significant effect on delaying decline of TSS (sugar receding) ( Figures 1A and S1). In the experiment of year 2020, 10 mg L −1 , 20 mg L −1 , and 50 mg L −1 GA 3 had no significant effect on the TSS content at ripening and harvest stages (96-110 DAF), but 50 mg L −1 GA 3 significantly reduced the degree of TSS decline at the later on-tree preservation stages (120 DAF, 11.96% higher; 130 DAF, 22.00% higher than the control) ( Figure S1A). Further, the results from year 2021 also showed that 50 mg L −1 GA 3 had a better effect than 10 and 150 mg L −1 GA 3 on delaying the decline of TSS at 105 to 130 DAF ( Figure S1B).
The 50 mg L −1 GA 3 treatment also showed an effect to maintain the content of antioxidants (mainly phenolics and flavonoids) and the activity of PAL, a key rate-limited enzyme for synthesis of phenolic acids and flavonoids during the later on-tree stages ( Figure 1B-D). The above-mentioned results of TFC and TPC might be explained by the pattern of changes in PAL activity, which was lower at 96 DAF but higher at 110, 120, and 130 DAF in the GA 3 -treated pulp compared to the control pulp ( Figure 1D). It was worth noting that the POD activity in the GA 3 -treated pulp was significantly higher at 96 and 110 DAF, but lower at 120 and 130 DAF ( Figure 1E). The Pearson correlation analysis indicated that TSS content showed a significant positive correlation with TFC (r 2 = 0.76, p = 0.0044) and TPC (r 2 = 0.7, p = 0.0077). A significant positive correlation was also found between TFC and TPC (r 2 = 0.60, p = 0.0395; Figure 1F). These results indicated that the GA 3 treatment significantly affected total soluble solids, antioxidants, and related enzyme activities in longan fruit during ripening and on-tree preservation. In recent years, preharvest or postharvest GAs application has attracted much attention in postharvest technology of fruits, due to their positive regulation of fruit size [13,[16][17][18][19][20], firmness [16,20], anthocyanin content [11], and PAL activity [11], as well as inhibition effect on respiration rate [15,20], postharvest ripening [15], and deterioration [9,15]. In this study, we found that preharvest GA 3 treatment with an appropriate concentration (50 mg L −1 ) led to decreased TPC, TFC, and PAL activity (similar to the results reported in the GA 3 -treated grape berry skin and flesh [13] and cherry [17]) in longan pulp at the mature stage. However, the treatment showed a significant effect to delay the sugar receding and maintain a higher sweetness (the most important commodity-related property of longan [3]). Moreover, the treatment enhanced the accumulation of antioxidants and PAL activity in pulp during the on-tree preservation.

Identification, Quantification, and Classification of Metabolites Detected in the GA3-Treated and Control Longan Pulp
In order to systematically reveal the effect of GA3 treatment on the metabolic profile of 'Shixia' longan pulp, a widely targeted metabolome by HPLC-ESI-triple quadrupolelinear ion trap (QTRAP)-MS was used to identify and quantify the metabolites. In total, 784 metabolites categorized into 14 classes (level 1) were detected in the 'Shixia' longan pulp (Figure 2A). The metabolome consisted of 46 alkaloids (5.87%), 110 amino acids and derivatives (14.03%), 159 flavonoid (20.28%), 23 lignans and coumarins (2.93%), 59 lipids database (Figure 2A). It was worth noting that more metabolites (784) were identified in this work compared to those in the previously reported analyses of 'Shixia' longan pulp (514 [2]; 706 [31]).    Figure 2D). At 96 DAF, more than 60% of the metabolites from the following 9 classes were up-regulated in the GA 3 (Figure S2A). At 110 DAF, 73.08% of tannins were up-regulated in the GA 3 -treated pulp ( Figure S2B). However, 65.38% of the tannins were down-regulated in the GA 3 -treated pulp at 130 DAF ( Figure S2C). The above results indicated that the secondary metabolic profile, especially the composition and content of antioxidants, in the on-tree preserved longan pulp was profoundly changed by the 50 mg L −1 GA 3 treatment. Although higher TFC and TPC were found in the 50 mg L −1 GA 3 -treated pulp at 110 DAF and 130 DAF, the effect of the treatment on the metabolic profile of longan pulp became weaker with the extension of on-tree preservation ( Figures 2C,D and S2B,C).  Figure 3D).

Significantly Differently Accumulated Metabolites between the GA 3 -Treated and Control Longan Pulp
A Venn analysis of DAMs with 1.5-FC from the GA 3 _96 vs. CK_96, GA 3 _110 vs. CK_110, and GA 3 _130 vs. CK_130 indicated that 25 metabolites-including 2 amino acids, 1 flavonoid, 1 coumarin, 3 lipids, 2 organic acids, 7 phenolic acids, and 8 tannins-were all significantly differently accumulated in the GA 3 -treated longan pulp at 96 DAF, 110 DAF, and 130 DAF compared to the control longan pulp ( Figure 3E,F). Moreover, 27 metabolites were found to be significantly differently accumulated in the GA 3 -treated longan pulp at both 96 DAF and 110 DAF compared to the control longan pulp; 6 metabolites were found to be significantly differently accumulated in the GA 3 -treated longan pulp at both 110 DAF and 130 DAF compared to the control longan pulp; and 38 metabolites were significantly differently accumulated in the GA 3 -treated longan pulp at both 96 DAF and 130 DAF compared to the control longan pulp ( Figure 3E,F). Therefore, 96 metabolites were found to be significantly differently accumulated in the GA 3 -treated longan pulp at 2 or 3 stages compared to the control longan pulp ( Figure 3E). A heatmap of these 96 common DAMs showed their expression pattern. The 96 common DAMs were composed of 6 alkaloids, 13 amino acids and derivatives, 17 flavonoid, 7 lignans and coumarins, 5 lipids, 7 nucleotides and derivatives, 8 organic acids, 1 other, 16 phenolic acids, 1 saccharide, 14 tannins. and 1 terpenoid ( Figure 3F). The above results indicated that the GA 3 treatment had a significant influence on the flavonoids (17/159), lignans and coumarins (7/23), phenolic acids (16/123), and tannins (14/26) which were related to antioxidant ability in the longan pulp during the on-tree preservation especially at 96 DAF and 110 DAF ( Figure S3). The above result was consistent with the reported positive effect of GA 3 treatment on the content of phenolics in jujube fruit [20], but contrary to the results of TFC and TPC reported in the grape berry flesh [13] and TPC reported in the cherry fruit [17]. Therefore, the influence of exogenous GA 3 treatment on the antioxidants depends on the fruit species, edible organ, treatment dose, time, and other factors. the influence of exogenous GA3 treatment on the antioxidants depends on the fruit species, edible organ, treatment dose, time, and other factors.

KEGG Enrichment of DAMs and Display of Important Metabolic Pathways
KEGG enrichment analysis showed that the DAMs from the GA3_96 vs. CK_96 were significantly enriched (FDR < 0.05) in 17 pathways, including arginine and proline metabolism; biosynthesis of alkaloids derived from histidine and purine; alanine, aspartate and

KEGG Enrichment of DAMs and Display of Important Metabolic Pathways
KEGG enrichment analysis showed that the DAMs from the GA 3 _96 vs. CK_96 were significantly enriched (FDR < 0.05) in 17 pathways, including arginine and proline metabolism; biosynthesis of alkaloids derived from histidine and purine; alanine, aspartate and glutamate metabolism; tyrosine metabolism; pyrimidine metabolism; phenylalanine metabolism; biosynthesis of alkaloids derived from ornithine, lysine and nicotinic acid; glycine, serine and threonine metabolism; biosynthesis of phenylpropanoids; phenyl-propanoid biosynthesis; biosynthesis of plant hormones; beta-alanine metabolism; flavone and flavonol biosynthesis; ABC transporters; galactose metabolism; butanoate metabolism; and purine metabolism (Figures 4A and S3). The DAMs from the GA 3 _110 vs. CK_110 were significantly enriched in five pathways, including arginine and proline metabolism, phenylpropanoid biosynthesis, β-alanine metabolism, and phenylalanine metabolism ( Figure 4B). Moreover, phenylpropanoid biosynthesis, linoleic acid metabolism, alanine, aspartate and glutamate metabolism, and phenylalanine metabolism were significantly enriched for the GA 3 _130 vs. CK _130 ( Figure 4C). It was interesting to note that the DAMs from the GA 3 _96 vs. CK_96, GA 3 _110 vs. CK_110, and GA 3 _130 vs. CK_130 were commonly enriched in phenylalanine metabolism and biosynthesis of phenylpropanoids. glutamate metabolism; tyrosine metabolism; pyrimidine metabolism; phenylalanine metabolism; biosynthesis of alkaloids derived from ornithine, lysine and nicotinic acid; glycine, serine and threonine metabolism; biosynthesis of phenylpropanoids; phenylpropanoid biosynthesis; biosynthesis of plant hormones; beta-alanine metabolism; flavone and flavonol biosynthesis; ABC transporters; galactose metabolism; butanoate metabolism; and purine metabolism (Figures 4A and S3). The DAMs from the GA3_110 vs. CK_110 were significantly enriched in five pathways, including arginine and proline metabolism, phenylpropanoid biosynthesis, β-alanine metabolism, and phenylalanine metabolism ( Figure 4B). Moreover, phenylpropanoid biosynthesis, linoleic acid metabolism, alanine, aspartate and glutamate metabolism, and phenylalanine metabolism were significantly enriched for the GA3_130 vs. CK_130 ( Figure 4C). It was interesting to note that the DAMs from the GA3_96 vs. CK_96, GA3_110 vs. CK_110, and GA3_130 vs. CK_130 were commonly enriched in phenylalanine metabolism and biosynthesis of phenylpropanoids.

The Effect of Preharvest 50 mg L −1 GA 3 Treatment on Storability and Postharvest Fruits' Antioxidants during Room-Temperature Storage
In addition to the reported effect of GA 3 on fruit quality (strawberry [11], 'French' prune [12], grape berry [13], pear fruit [14], cherry [17], sour cherry [18], "Washington Navel" orange [19], jujube fruit [20]) at harvest, as well as the maintenance of VC and phenolic substances after harvest, preharvest treatments with oxalic acid (on sweet cherry [32] and lemon fruit [33]), melatonin (on pomegranate [34]), and methyl jasmonate (on plum [35]) were proven to improve fruit quality and storability, stimulate the antioxidant system, and enhance accumulation of bioactive compounds. In this study, the longan fruits for the storage experiment were sprayed with 50 mg L −1 GA 3 or the control solution 2 times (at 85 DAF and 95 DAF) and harvested at 105 DAF. Our results showed that 50 mg L −1 GA 3 significantly increased the storability of 'Shixia' longan fruit by delaying mass loss, pericarp browning, increase in relative conductivity, and aril breakdown during the RT storage ( Figure 5A-D). As shown in Figure 5A, lower total mass loss (%) was found in the treated longan fruit throughout the storage (significant at 2 DAH and 4 DAH). More importantly, significantly delayed development of pericarp browning and lower relative conductivity (significant at 0 DAH to 6 DAH) were observed in the treated longan fruit ( Figure 5B,C). In addition, the aril breakdown index of the treated longan was significantly lower than that of the control, which developed quickly at the two later stages (8 DAH and 10 DAH) ( Figure 5D). However, the 50 mg L −1 GA 3 treatment did not significantly influence the TSS and TA content of 'Shixia' longan at harvest. No significant difference in TSS and TA content was observed between the treated longan and the control throughout storage ( Figure 5E,F). The above results indicated that the 50 mg L −1 GA 3 treatment significantly improved the storability of 'Shixia' longan fruit. In addition to the reported effect of GA3 on fruit quality (strawberry [11], 'French' prune [12], grape berry [13], pear fruit [14], cherry [17], sour cherry [18], "Washington Navel" orange [19], jujube fruit [20]) at harvest, as well as the maintenance of VC and phenolic substances after harvest, preharvest treatments with oxalic acid (on sweet cherry [32] and lemon fruit [33]), melatonin (on pomegranate [34]), and methyl jasmonate (on plum [35]) were proven to improve fruit quality and storability, stimulate the antioxidant system, and enhance accumulation of bioactive compounds. In this study, the longan fruits for the storage experiment were sprayed with 50 mg L −1 GA3 or the control solution 2 times (at 85 DAF and 95 DAF) and harvested at 105 DAF. Our results showed that 50 mg L −1 GA3 significantly increased the storability of 'Shixia' longan fruit by delaying mass loss, pericarp browning, increase in relative conductivity, and aril breakdown during the RT storage ( Figure 5A-D). As shown in Figure 5A, lower total mass loss (%) was found in the treated longan fruit throughout the storage (significant at 2 DAH and 4 DAH). More importantly, significantly delayed development of pericarp browning and lower relative conductivity (significant at 0 DAH to 6 DAH) were observed in the treated longan fruit ( Figure 5B,C). In addition, the aril breakdown index of the treated longan was significantly lower than that of the control, which developed quickly at the two later stages (8 DAH and 10 DAH) ( Figure 5D). However, the 50 mg L −1 GA3 treatment did not significantly influence the TSS and TA content of 'Shixia' longan at harvest. No significant difference in TSS and TA content was observed between the treated longan and the control throughout storage ( Figure 5E,F). The above results indicated that the 50 mg L −1 GA3 treatment significantly improved the storability of 'Shixia' longan fruit. Similar effects as reported in previous works [32][33][34][35] were also observed in the GA3 treatment on longan samples: significantly higher VC content was detected in the treated longan pulp at 2, 4, and 10 DAH ( Figure 6A). Lower TFC at the earlier stages of storage (2 Similar effects as reported in previous works [32][33][34][35] were also observed in the GA 3 treatment on longan samples: significantly higher VC content was detected in the treated longan pulp at 2, 4, and 10 DAH ( Figure 6A). Lower TFC at the earlier stages of storage (2 DAH, 4 DAH, and 6 DAH) but significantly higher TFC at 8 DAH and 10 DAH were observed in the treated longan pulp compared to that in the control ( Figure 6B). The TPC slowly increased along with the RT storage time and showed no difference between the treated longan pulp and the control pulp ( Figure 6C). Interestingly, the GSH content in the treated longan pulp increased and was significantly higher than that in control pulp during the entirety of RT storage (2 DAH to 10 DAH) ( Figure 6D). DAH, 4 DAH, and 6 DAH) but significantly higher TFC at 8 DAH and 10 DAH were observed in the treated longan pulp compared to that in the control ( Figure 6B). The TPC slowly increased along with the RT storage time and showed no difference between the treated longan pulp and the control pulp ( Figure 6C). Interestingly, the GSH content in the treated longan pulp increased and was significantly higher than that in control pulp during the entirety of RT storage (2 DAH to 10 DAH) ( Figure 6D). The VC content in the control longan pericarp increased from 0 DAH to 6 DAH, and then slowly decreased. Significantly higher VC was observed at 4 DAH and 10 DAH in the GA3-treated pericarp ( Figure 6E). The TFC in the control longan pericarp was significantly higher than that in the treated pericarp at harvest (0 DAH), but it decreased rapidly from 4 DAH to 10 DAH. However, the TFC in the treated longan pericarp was significantly higher than that in the control at the later storage stages (4, 8, and 10 DAH) ( Figure 6F). Similarly, the TPC in the treated pericarp was significantly lower than that in the control pericarp at the earlier stages of storage (0 DAH and 2 DAH), but it was significantly higher The VC content in the control longan pericarp increased from 0 DAH to 6 DAH, and then slowly decreased. Significantly higher VC was observed at 4 DAH and 10 DAH in the GA 3 -treated pericarp ( Figure 6E). The TFC in the control longan pericarp was significantly higher than that in the treated pericarp at harvest (0 DAH), but it decreased rapidly from 4 DAH to 10 DAH. However, the TFC in the treated longan pericarp was significantly higher than that in the control at the later storage stages (4, 8, and 10 DAH) ( Figure 6F). Similarly, the TPC in the treated pericarp was significantly lower than that in the control pericarp at the earlier stages of storage (0 DAH and 2 DAH), but it was significantly higher than that in the control at 4, 8, and 10 DAH ( Figure 6G). Moreover, the GSH content in the control longan pericarp increased from 0 to 4 DAH, but then decreased. Lower GSH content was observed in the treated longan pericarp at only 2 and 8 DAH ( Figure 6H). In total, the 50 mg L −1 GA 3 treatment significantly enhanced the accumulation of VC in longan pericarp and pulp and GSH in pulp at the later storage stages, or helped maintain the TFC in pulp and both the TPC and TFC in pericarp. A similar effect was reported in the literature, that the combined treatments of CPPU and GA 3 on green mature banana significantly delayed the loss of ascorbic acid and total phenols, retarding chlorophyll degradation during storage [15].
Many previous works indicated that longan pericarp browning was related to phenolic metabolism [36], and the maintenance of flavonoids, phenolics, and other antioxidants (i.e., VC and GSH) helped balance the scavenging ability and production of oxygen free radicals [26,30,36], which protected cytomembrane integrity [7] and, thus, inhibited pericarp browning. Flavonoids have been proven as essential barriers against pathogen attack [37]. In addition, higher levels of VC, flavonoids, GSH, and total phenolics, as well as higher ability of scavenging free radicals, have been proven to help maintain the structural integrity of longan pulp cell membrane and suppress the pulp breakdown occurrence in postharvest longan fruit [38]. Our results indicated that the 50 mg L −1 GA 3 treatment significantly improved the storability of 'Shixia' longan fruit, possibly through maintaining or increasing the antioxidant content at the later storage stages but not at harvest.

Conclusions
Preharvest GA 3 treatment with an appropriate concentration (50 mg L −1 ) showed a significant effect in delaying the sugar receding of 'Shixia' longan, a very important main variety in China. Moreover, higher content of antioxidant bioactive compounds (mainly phenolic acids, flavonoids, tannins, lignans, and coumarins) and higher PAL activity were found in the GA 3 -treated pulp during the on-tree preservation stages. More importantly, 50 mg L −1 GA 3 did not influence the TSS and TA content, but significantly delayed pericarp browning and aril breakdown, which resulted in lower pericarp relative conductivity and mass loss but higher vitamin C and TFC, along with reduced glutathione in pulp and higher vitamin C, TFC, and TPC in pericarp at the later RT storage stages. Overall, these results confirmed that preharvest 50 mg L −1 GA 3 spraying could be a practical tool to delay the sugar receding of longan fruit, and thus prolong its harvest time and supply time as fresh fruit, as well as enhance antioxidant compounds, improve functional properties, and reduce postharvest loss during RT storage.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/foods12102032/s1, Figure S1. The TSS content of the GA 3 -treated and control longan pulp in the year 2020 and 2021; Figure S2. Statistics of metabolites in each class according to the fold change by the GA 3 vs. control longan pulp; Figure S3. The fold changes of the metabolites from the important pathways. CK: Control; GA 3 : 50 mg L -1 GA 3 .